app.py

import streamlit as st
from transformers import pipeline

sentiment_analysis = pipeline("sentiment-analysis", model="nlptown/bert-base-multilingual-uncased-sentiment")

def convert_label_to_sentiment(label):
    if label in ["4", "5"]:
        return "Positive"
    elif label in ["1", "2"]:
        return "Negative"
    else:
        return "Neutral"

def perform_sentiment_analysis(text):
    result = sentiment_analysis(text)
    return {'label': result[0]['label'], 'score': result[0]['score']}

def main():
    st.title("Financial Sentiment Analysis")

    # Input for financial content
    financial_content = st.text_area("Enter Financial Content:", "With the launch of Apple Silicon, Apple shares have increased")

    # Perform sentiment analysis on button click
    if st.button("Submit"):
        if financial_content.strip():
            sentiment_result = perform_sentiment_analysis(financial_content)
            sentiment = convert_label_to_sentiment(sentiment_result['label'][0])
            st.success(f"Sentiment: {sentiment} | Score: {sentiment_result['score']:.2f}")
        else:
            st.warning("Please enter financial content before submitting.")

if __name__ == "__main__":
    main()